The clang backend's guard retry still maps an aggregate line to a header (2026-10-03)

Index: Known gaps

Status (re-verified 2026-10-09 at 456989f): OPEN. abicheck/extract/headers/clang/error_header_retry.py:69-79 still maps the aggregate line index to the header. The entry notes it is latent: not broken unless a preamble is added. The writers are now in extract/headers/clang/backend.py:304 and clang_layout_tool.py:203.

extract/unparseable_header_fallback.py attributed a castxml diagnostic to header N-1 from the aggregate frame's line N. When the castxml aggregate gained a preamble include on its first line, that rule dropped the healthy neighbour of a rejected header and the directory dump failed. It now names the input by the file the aggregate frame includes, which does not depend on the aggregate's layout (bug class extraction.aggregate_layout_inverted_by_line).

The clang backend's direct-inclusion-guard retry (extract/headers/clang/error_header_retry.py, _headers_failing_in_aggregate) still uses the line rule. It is correct today: both of its writers (dumper.py's _write_agg, clang_layout_tool.py's _write_agg) put one include per line with nothing before them. A preamble added there would reintroduce the defect; attributing by the included file, as the castxml fallback now does, closes it. Not changed here because nothing is broken and the function's test suite encodes the line layout in every case.